INDEX

Explanations

Levels and Stages

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 París

-0.48

 trattano

-0.46

agisse

-0.44

 getArguments

-0.44

copg

-0.43

Abuja

-0.43

 Viena

-0.43

-0.42

Abril

-0.41

łoży

-0.41

POSITIVE LOGITS

AndEndTag

0.72

+#+#

0.64

цездатний

0.62

 Ause

0.60

 ويكيپيديا

0.60

entuh

0.56

borderSide

0.56

 EconPapers

0.53

imine

0.53

RunAsync

0.52

Activations Density 0.003%